How to Start a Computer Games Company

Picture of Matthew Stibbe
Posted by Matthew Stibbe

This article is about the hard reality of starting and running a computer games company. I started Intelligent Games (IG) in 1988 because I have always loved computer games and I hired my first employee and moved into real offices in 1993. I sold the company and left in 2000. In that time we released sixteen titles, which together sold over two million copies. The company grew to more than sixty-five people and worked with some of the biggest publishers in the business: EA Sports, Westwood Studios, LEGO and Hasbro. We got there without debt and without external investment. It is from this experience that I am writing this article – as they say in America: ‘your mileage may differ’.

(I wrote this article for the law firm Osborne Clarke for their website in 2001 based on speeches I gave to the Game Developers’ Conference in 1998 and 1999. It isn’t completely up-to-date and the market has changed considerably even in the last couple of years. As I update and re-read this article in 2019, it all seems a very long time ago.)

Starting a games company isn’t easy

This article is intended to give you some insights into the life of a developer: some of the unspoken truths and hidden history as well as outlining some of the things that can help you start a development company. In particular, I am going to give some advice about generating ideas, sales and marketing and the boring (but vital) bits: management, offices, cash flow, and funding. Osborne Clark can advise you on the legal side of the business, so I won’t talk about that at all. Finally, I will speculate on some of the things that build value in a development company and possible exit routes for you once you are up and running.

The Unspoken Truths

I don’t want anyone reading this article and thinking that starting a game development company is an easy and fast way to make loads of money. It isn’t. Therefore, I am starting with a list of the unspoken truths of the industry – the things they don’t tell you in Edge magazine!

There are over 4000 games titles published worldwide each year. 50% of them sell less than 10,000 copies. Only the top couple of dozen sell enough to be classed as hits. Typically, this is in the 500,000-1,000,000 unit sales range on PC and two to three times that on Playstation One.

A hit game is a wonderful thing – it will make money and establish your reputation. The flip side is that the ratio of hits to flops is appalling. Further, there are hundreds of third party developers out there so there is a lot of competition.

Finally, if you analyse the hits in any given year over two-thirds will be sequels of existing titles (a franchise hit) or based on a massive non-games brands such as LEGO or Barbie. Publishers are reluctant to let untested developers loose on valuable brands or their precious franchises. The number of original titles from independent studios that make hits is very small indeed. Not only does this make publishers wary but also it dramatically decreases the odds against your project ever earning royalties.

Marketing Strategy foundations course CTA

The dirt on royalties

On the subject of royalties – most publishers talk as if they are almost guaranteed when they are negotiating a deal. Don’t believe it. Generally, royalty rates are calculated so that the product has to exceed its sales targets just to break even on the advances. Some people are able to get fantastic royalty rates that buck this tendency but usually this is because either they have had a previous hit or because they are part- or wholly-funding the game development themselves and taking the risk away from the publisher.

Why your game needs a brand

I gave a presentation on working with brands at the Game Developers Conference in San Jose in March 2000. In my opinion, working with or creating strong brands is absolutely central to success as a developer. You can read that on my website at www.stibbe.net/brands_speech.htm. I have lots of evidence for my view but there are several well-known who eschew brands and develop original titles.

Games are getting bigger and more expensive. My first game in 1988 cost £35,000 to develop and took two people about nine months. At IG, we have had several projects with budgets over £1m and almost all the rest are over £500,000. With the exception of kids titles, most of these games take 12-18 months with teams of at least ten people to complete. At the extreme end of the spectrum, the FIFA Soccer team is about 120 people and they produce one title a year. In other words, in ten years teams and budgets have increased by a factor of ten. This is making it increasingly difficult for small teams to appear credible to large publishers.

Games get killed. Deal with it.

A final unspoken truth is that publishers deliberately kill at least 50% of all the titles they sign before completion. In IG’s twelve-year experience we do two paid-for designs for every one that gets put into full production. Half our in-production titles get killed. This is a fact of life for all developers but few like to admit it because it looks like admitting weakness. The reasons are many:

  • Changing markets
  • Publishers going bust or retrenching
  • Development delays
  • The release of an unexpected competitor product
  • Occasionally sheer publisher incompetence.

Having a product killed is extremely uncomfortable. Even with a cancellation fee, you have to rapidly find work for the team by selling a new project and deal with your nicely planned world being turned upside down and you have to deal with the team’s feelings about this too. Usually a termination is preceded by an uncomfortable period where publishers behave strangely and sometimes stop paying. This can serve as a warning but can also be very stressful.

What games publishers are really like

I may seem very down on publishers. My honest view is that in the majority of cases they are like dinosaurs in Jurassic Park. They aren’t evil – they just follow their own genetic behaviour and pursue their own interests instinctively. If signing your game will help them, they’ll sign it. If killing it will help them, they’ll do that. It is, however, important to remember that they are not monolithic organizations that act coherently and rationally.

Producers control the flow of information into and out of a publisher and they are human beings (most of them) and they have bosses who have bosses. Therefore, decisions may not be made on the best information or in a completely open, rational environment. This isn’t malice, it’s the human condition. If you are aware of these problems from the start then you have a better chance of pre-empting them. As with everything, of course, there are exceptions.

  • Some of the nice things that publishers have done for IG over the years include:
  • Paying for US mobile phones and trips
  • Paying invoices very quickly or even ahead of approval to help our cash flow
  • Giving us end-user information and research
  • Helped with recruitment
  • Provided tools and code, sound and video drivers
  • Some of them cancelled games in an open, honest and decent way and paid a fair cancellation fee
  • Provided a kind of support network for me personally

Finally, I cannot forget that publishers’ advances built the company.

The secret history of my computer games company

We have published a lot of games, but we’ve had as many canned. Nobody saw Dark Hermetic Order, Flying Circus, Bloodline, Conjure, King of Wall Street, Deadline News, Cops and Robbers. Even though all these titles were green lit and contracted for. Why? Because they were killed by publishers for reasons largely outside our control.

Our last batch of products: Dune, Loco, Action Man and the current crop in development were mostly on time. The first four published games, however, were between six and fourteen months late. Since we were working on fixed-price contracts this cost us dear. We learned some enormously valuable lessons about development, code reuse, project management, personnel management, testing, scheduling, publisher relations and deal making from these disasters that have helped the company improve dramatically. However, it took us something like four years to learn these lessons and put them into practice.

One publisher went bankrupt on us and two others completely pulled out of publishing in the middle of our development.

Finally, only two of the first ten people I hired are still with the company. Actually, I think this is pretty good compared to the norm. In the context of starting up a business, this says that it is hard to find and retain the right staff.

Why do it?

If you’re still reading this, I haven’t been able to put you off! So why set up an independent development company?

For me the main reason was to make the games I enjoyed playing. I loved Harpoon and SimCity and so the first two games I designed were USS Ticonderoga and SimIsle. Happily, I was able to sell them to 360 Pacific and Maxis while working from home with only one employee. I think this joy of creation is the most frequently cited reason for people setting up on their own. Another important element is the chance to be master of your own destiny.

Lots of people would love to be their own boss. As one wit put it “everyone wants to go to heaven but no-one wants to die.” I would imagine that working in the games industry in some companies would be an affront to one’s integrity and setting up on your own would be an opportunity to do things right.

There can be an element of status and ambition to it – being the creator of a hit game puts you in the spotlight in this industry like no other. There is always the promise of lavish rewards if you are successful. All these things do happen. Finally, if you are highly stress-motivated and really enjoy working long hours then this is definitely the job for you.

Games company startup checklist

There is lots of information to help you form a company and manage the practical side of starting you own business. Amazon or your local bookshops will have useful books of advice. Also, your bank will probably have lots of materials for people setting up a business. This check list isn’t exhaustive but it will give you an idea, in advance, of the things you will need to do:

Find a good lawyer and a good accountant. They will understand that you are starting out and should tailor their services (and fees) to your needs but they won’t be free. You should work with them to fix prices and services in advance so that there are no surprises.

Practical tips for games company startups

  • Open a separate business bank account
  • Contact the Inland Revenue and get their new employer’s pack
  • You should probably consider appointing a payroll agency
  • There is a turnover threshold at which you have to charge VAT (but you can reclaim VAT on your expenditures). Contact HM Cusoms and Excise for their starters pack for more information.
  • You will need to decide whether to be sole trader, a partnership or form a limited company. Generally, it is probably most sensible to be a limited company. 
  • While you can start up in your garage (or in my case, in the kitchen) you will soon want an office. I have found it very helpful to use an agent to find an office. While you will need to pay for this, they will save you time and cut a better deal for you. We used Stanford Webster for our last office and they did a fine job for us.

On a personal level, you will probably need to warn your friends and family that you are starting up a business and that it will be hard work and stressful. Having a personal support network is vital. My experience is that once people start seeing you as a ‘boss’ they find it difficult to be completely free and friendly. You may not get the support and camaraderie as an owner that you are used to as a colleague. Of course, this may be just me!

Marketing your games business

There’s an old joke: if you take someone out for dinner and at the end of the meal you say “I’m really good at sex, do you want to come back to my place” that’s advertising. If you say “you look lonely, perhaps you’d like to stay the night with me” that’s marketing. If they say “I’ve heard you’re a fantastic lover, can I come back to your place”, that’s public relations. Marketing is all about working out what you can offer and what prospective clients need.

The first thing to do is to identify what type of product you want to develop and make sure that there is a market for it. Work out what genre and platform you want to work in – play to your strengths and try to consider several options. Do this before you start spending time actually developing game ideas so that you don’t waste time designing games in a genre that you subsequently reject.

Be selective, be passionate

You should make your choices based on what you feel passionate about but you should also be realistic about what will sell. Once you have settled on a genre, get creative and come up with as many ideas as you possibly can. Be prepared to reject 80% of the ideas you come up with and so don’t spend too much effort on any single idea at this stage.

At IG we tend to do a one page ‘concept brief’ to describe a game. It consists of the working title, a one sentence description of the game’s ‘high concept’ (for example the high concept of the film Aliens was ‘Jaws in space’), a description of the key features and USPs (unique selling propositions) and a description of the game play and visualisation as well as practical information about target platform and target audience.

Testing your game ideas

There are two exercises you can do to test the validity of an idea at the one page point. First you can review as many prospective competitors to see how well they reviewed, how well they sold and what you’re game offers that they didn’t. If you can’t offer consumers something new or exploit a genuine gap in the market then you should move onto a new idea. Equally if no games like yours sold more than 50,000 units or there is an indestructible market leader, then move on. The second test is to see if you can enthuse someone else (who likes games but not necessarily a publisher) about your idea and see if they ‘get it’.

Before they can go to a prospective client, these ideas need to be well presented. Avoid spelling mistakes and bad layout – what you present to someone is like an advert for your style and approach as well as the product. If it looks slapdash that’s the image your company will have. Try to make sure the ideas are technically feasible and not overly ambitious, especially considering the resources you have available to develop the product.

Build your database

At this stage you need to target possible publishers. We do this by maintaining a database of publishers with as much information about them as we can find and lists of their current products. This lets us see whether they have a gap or requirement for the products and services we have to sell. Doing this kind of market research isn’t difficult. We use this database to record contact reports after each meeting and to log action items such as things we have promised to send. The other thing we do was to keep up a fairly constant dialogue with people at shows, by email and in face-to-face meetings. It is easier to do this when you are more established, but it is worth cold calling into as many publishers as possible and asking the people you meet what they need and what they are looking for.

There is another way to figure out what products to do and that is to approach publishers with your development resources and ambitions and see if they have any work to give you. Most of our large contracts have come from building a relationship with a publisher over small data disk type projects. For example before we signed Dune 2000 with Westwood Studios we did two data disks for them. Taking this approach requires that you have some capabilities or resources to sell first.

Selling your game idea

Once you have something to sell, you’ll need to start knocking on doors. It will help if you already have contacts in the industry or some industry track record (for example if you’ve left an established firm to start up your own business). As well as selling yourself, your company and your ideas you also need to be evaluating prospective clients. I have put together a publisher purity test based on my experience to give some kind of benchmark on the performance of a publisher during the sales process.

How to evaluate different publishers

It’s not black and white, but good publishers consistently score highly on this test and bad publishers consistently score low!

  • Do they return calls? How often?
  • Do you meet different people? Senior people?
  • How long does it take to do the contract paperwork?
  • Are there promises and actions consistent?
  • Is there a good cultural fit?
  • Do they work with same developers again?
  • How many projects do they kill? Why?
  • Are they suing any developers? Why?
  • Do you personally like or respect their current range of products

Give good meeting

Some good questions to ask at a first meeting:

  • How do you do distribution in each of your territories (especially the US)?
  • What is the proportion of internal vs. external development?
  • How many developers have you sued?
  • What proportion of your developers do more than one release with you
  • How many products earn back their advances and earn the developers royalties?
  • For what reasons have you killed products in the last year?
  • Are you profitable? (This is a killer for most publishers!)
  • What is the approval process for getting a game into production?
  • How long does it take your legal department to issue contract once a deal is green lit? [Anything more than a month is pretty poor]
  • What are the three worst-performing games in your catalogue? Why?
  • Ditto for the three best-performing games.

How to make a good impression

You will evolve your own sales style but here are few tips.

  • Be honest but positive about the product and your capabilities.
  • Don’t hide difficulties – be up front about them but make sure you also say what your plans are for dealing with them.
  • Try to find and talk to key people who are part of the actual green light process.
  • Understand their process: who signs off, how long does it take, when does it happen etc. The person you are talking to has a job and objectives like you and that much of selling is about personal relationships and mutual confidence.
  • Keep the ball in their court so don’t keep them waiting for documents or other things that you’ve promised.
  • Be courteous (to everyone, receptionists included: years ago I once visited Virgin Interactive and saw Richard Branson on reception for two hours!) but firm. I once wanted to sell a game to a publisher that simply wouldn’t accept unsolicited proposals or calls so I went out to America and sat in their reception for five hours until eventually the receptionist went and found someone for me to talk to. I was polite but firm! Luckily, I haven’t had to do that often.
  • Finally, be humble and prepared for rejection and remember there’s always a next time.

Expect the worst, hope for the best

In my experience:

  • They probably won’t call you back
  • It always takes longer
  • You’re lucky if they remember to say ‘no’
  • The producer always changes

The publishers who pass the purity test and who give good answers to the list of questions will probably be exceptions.

Get legal advice

You should probably have a lawyer’s advice or that of a good agent when you start approaching publishers. I made the mistake very early on of signing a design letter of intent which ended up giving away a lot of rights to a publisher before they gave me a full production contract. On the other hand, it is worth remember that ideas are very cheap and that in twelve years’ experience I have never seen one of IG’s ideas ripped off by a publisher so don’t be precious about confidentiality agreements and things of that ilk. Of course, you should take your own counsel on this.

My final bit of advice on the sales side is ‘prepare prepare prepare’. One thing that you probably have a lot of is time and this is the one commodity that publishers are short of. For example, I would count on putting in three to four times the preparation time for a negotiation as the producer on the other side of the table.

Do you need a game agent?

In my experience you still end up doing the same amount of work on sales and marketing when you appoint an agent but you benefit from their range of contacts, industry experience and negotiating skill. IG used an agent on only one deal we did. The bad news is that you will end up giving them 10-12.5% of your gross receipts (i.e. of money that comes in the door). The good news is that they may well get a better deal for you and you only pay them if they succeed.

This is a list of four UK agencies that I am aware of:

  • Mr. Mark Cochrane (The Games Agency): mark@thegamesagency.com
  • Ms. Jacqui Lyons (Marjacq): jacqui@marjacq.com
  • Mr. Kirk Owen (Octagon): submissions@octagon1.com
  • Mr. John Cook (Bad Management): badman@cix.compulink.co.uk

Pricing and scheduling game development

Most publishers will want you to sign a fixed price contract where you deliver against a pre-agreed specification and schedule. If you get the calculations wrong and need more people or more time to deliver the agreed features it is your risk. This means that great care must be taken in the scheduling, specification and pricing of the product.

Code complete

The best textbook for project analysis and design is “Code Complete” from the Microsoft Press. IG uses Microsoft Project for scheduling although we have evolved a fairly sophisticated and roundabout way of using it that works around its problems and fits in with the way we work. This is a trade secret and you’re going to have wrestle with Project on your own! It is important to be as detailed, specific and realistic as possible in building a schedule.

Breaking down schedules into manageable units

IG tends to work to 2-5 days tasks in building its schedules and gives individuals responsibility for estimating their chunks of work. We assume that people will work four-day weeks actually coding. The other day a week takes account of holidays, sickness, meetings, paperwork etc. Our experience is that this is pretty accurate even when people are actually working 50-60 hours a week (or more in the last couple of months). If you overschedule, say at six working days a week, then you had better be absolutely sure that everyone will sustain that level of effort actually developing the game for the entirety of the project. It is an easy mistake to make.

Budgeting

You should budget your projects on the assumption that you won’t receive royalties and you should aim to recover all your costs, including a fair proportion of your overheads, from the project advances. We calculate budgets based on the schedule by multiplying the number of developer months required to complete the project by an average cost per developer month plus an allocation of overhead.

As a rule of thumb budget £1 for overheads for each £1 of salary cost. Since publishers like to backload a payment schedule so that up to 30% comes in the last couple of months, you need to make sure that the payment schedule covers your month-to-month cash flow needs. It isn’t unreasonable to ask for a 10-20% contingency margin over and above the development costs but don’t ask for an out and out profit if the project has a royalty component.

Don’t forget to budget for console development kits (Playstation 1 kits are £12,500), sound, voice recording and music (a typical game will have a sound budget in the £30-50k range).

Management

I gave a speech about managing creativity that is also on my website so I won’t bore you by repeating it here. The only other advice I have is to read “PeopleWare” by DeMarco – it’s a quick read but invaluable.

Finances

In a large development company a typical developer month costs about £6-8k including a share of the overheads. A startup can have smaller overheads. Even so, you should expect to budget a minimum of £3-4k a developer month as a rough rule of thumb. Once you have a development contract you can expect a publisher to cover these costs by way of advances against future royalties. Until you have a contract – and it can take three to six months – you will need to cover these costs yourself. There are several sources of startup capital:

  • Your own money (if you have it), money from friends and family (if they have it) or a loan or second mortgage
  • Business angel funding or venture capital (there is a lot of information on the British Venture Capital Association on both sources at www.bvca.co.uk).
  • A second job – I started IG with money I earned as a contract programmer.

It isn’t easy to raise money for a complete startup but it is possible. Above all you’ll need a good business plan. There are many good books about writing a business plan. There are a couple of good websites called www.startups.com and www.startups.co.uk that have a lot of useful information about writing a business plan. This is another area where banks and accountants can be very helpful.

Budget figures for overheads

It is very hard to give definite figures for the costs of setting up a business, but I have tried to set out below some reasonable guesstimates for the annual costs – assuming a London base – of the various elements to give some idea of what to budget for.

  • Office space: 150 square feet per person at £15-20 per square foot per year plus £3-5 per square foot service charge plus £5-7 per square foot in rates. 
  • Phone system: £2000 + £150 per extension plus line rental and call costs
  • Internet connection: £1500 connection and £5000 p.a. for a reasonable leased line
  • Servers: File server, mail server: about £2000 each plus £2500 for hubs, cabling and so on for a network
  • PCs: £1500-2000 each
  • Software: £3000 per artist (3DS Max and Adobe Photoshop), £500ish per programmer (Microsoft Visual C++, Visual Sourcesafe).
  • Network software: Exchange and NT have per user licences of about £50 each per person
  • Audit and accountancy fees: £5000
  • Legal advice: £3000 (depends on how much you use them for, of course)
  • Insurance: £5000+ (Office and contents, public and employers liability, professional indemnity)
  • Stationery, consumables, post and packaging

Programmers and artists earn between £18,000 and £35,000 depending on experience and qualifications. There isn’t always a good correlation between cost and competence, however, and you need to interview and test prospective employees carefully. Developers with management responsibilities can expect to earn more than this if they’re good and experienced.

Cash flow management in games companies

One of the toughest transitions for someone new to running a business is dealing with cash flow and credit control. When someone else is responsible for paying or collecting the bills, its easy to forget how important it is. Since the outgoing costs of a business, especially the wages, go out at fixed times each month but income is variable and depends on the timeliness of deliverables, client approvals and the time it takes for a client to actually pay up there is a lot of tension in a small business between cash in and cash out.

Typically a large publisher will take two to four weeks to review a milestone and assuming they approve it another four weeks to actually send you the money. This means that you will be working, effectively, three months in arrears the whole time. This can be partly compensated for by getting a large signing fee when a contract is signed and also by managing credit control and approvals attentively.

How to get paid on time

My advice on this point is to watch the cash flow and credit control carefully. Poor cash flow kills more games companies than any other problem. Do a week by week cash flow plan and updated it weekly. Track the dates on which you send milestones, invoices and so on. Try to understand a publisher’s approval and payment cycle. Usually a milestone needs to be signed off by two or three people before it can be passed to accounts for payment. If you know who these people are it is easier to chase them.

You should make your needs and expectations clear to everyone in the chain at the beginning and if you have problems you should explain them clearly to the senior people who can help sort them out. It’s difficult to avoid but don’t lose your cool or bullshit anyone.

Dealing with your suppliers

Equally if you have to hold payment to people you owe money too, be honest and explain the reasons and tell them when you think you can pay them. As soon as you can afford it you should aim to have a book keeper to help with this and then later an in-house accountant. From the beginning, you should aim to have a monthly profit and loss statement, balance sheet, cash flow project and aged debtors and creditors analysis. Most small developers go bust and most small developers don’t run their businesses like business – go figure. As someone once said to me: “there’s plenty of time to do nothing when you’re dead”.

The Institute of Directors do a very good ‘accounting for non-financial directors’ course that is open to non-members. This three day course will give you a good introduction to the interplay between cash flow, profit and loss and the balance sheet and other arcane accounting arts.

Building value in your games company

In the long run, there are several objectives open to you for your business:

  • Triumphant independence – profits, royalties, critical success and continued growth
  • A trade sale to a publisher
  • A partial trade sale plus a multiple-product deal with a publisher
  • A stock market floatation
  • Transition to becoming an affiliate label or publisher

Some of these routes will require injections of funds from venture capitalists or other sources of funds. Other routes let you take some money out of the business. There seems to be a natural limit to the size of truly independent development companies and it is at about 100 people. Beyond this size they tend to stop being independent in some way. Of course, the other route is failure and this is, sadly, the most common outcome. Something like 25% of all games development companies in the UK fail each year.

How to be a successful game developer

The good news is that good development companies can be worth two, three, four, even ten or more times their turnover to a prospective buyer. This is because good development talent and consistent hit-producers are very scarce. The basic value of a development business is somewhere between its break-up value and the cost of reconstructing it from scratch – usually less than a multiple of one times turnover. To drive this number up, you need:

  • Hits
  • Reputation
  • Close relationships with a publisher
  • Valuable proprietary technology or brands
  • Synergy with a particular buyer
  • An inside track on some new technology (like the Internet or next-gen consoles)
  • Famous individuals on staff

So, should you start a games business?

Overall, the whole prospect must sound pretty scary but…

  • People have done it and succeeded
  • There are decent, effective publishers out there
  • It is possible to learn from mistakes
  • The market is the same for everyone
  • The market is growing overall
  • Some people make tons of money out of games

In conclusion, I am reminded most of the strapline for the film “The Commitments” which read “they had nothing to lose but they were prepared to risk it all.” If you make this choice, I wish you every success.

How to choose an agency for your new website